T-Shirt Sizing vs Story Points: Which to Use?
Verdict: T-shirt sizing (XS, S, M, L, XL) is simpler and works best for early, rough estimation when items aren't yet well understood — often used during initial backlog triage or roadmap-level planning. Story points offer more precision and are better suited for sprint-level planning once items are refined enough for the team to reason about relative size more specifically. Many teams use both, at different stages of the same backlog item's lifecycle.
Quick facts
- T-shirt sizing uses broad categories; story points use a numeric scale (often Fibonacci-like) for finer relative comparison.
- T-shirt sizing is faster and requires less precision, making it well suited for early-stage or less-refined items.
- This connects to Story Points vs Hours Estimation, covering the related comparison between points and literal time estimates.
Side-by-side comparison
| T-Shirt Sizing | Story Points | |
|---|---|---|
| Scale | XS, S, M, L, XL (a handful of broad categories) | Numeric, often Fibonacci-like (1, 2, 3, 5, 8, 13...) |
| Precision | Lower — broad relative buckets | Higher — finer-grained relative comparison |
| Speed | Faster, since fewer distinctions to debate | Slower, since more granular discussion is often needed |
| Best used for | Early-stage roadmap or backlog triage, before items are well understood | Sprint-level planning, once items are refined |
| Team velocity tracking | Not typically used for this purpose | Core input for calculating team velocity |
Why T-shirt sizing works well early
When a backlog item is still rough — not yet refined with clear acceptance criteria or a solid understanding of the work involved — trying to estimate it precisely with story points can create a false sense of accuracy the team doesn't actually have. T-shirt sizing's broader categories match this genuine uncertainty better, letting the team quickly triage a large batch of rough ideas ("this feels like an L, this feels like an S") without getting bogged down in a level of precision the information doesn't yet support.
Why story points work better for sprint planning
Once a backlog item has been refined — clear acceptance criteria, understood scope, no major open questions (see Definition of Ready) — the team has enough real information to reason about relative size more precisely. Story points, particularly with a Fibonacci-like scale, let a team distinguish between meaningfully different sizes (a "3" versus a "5") in a way that supports reliable sprint planning and, over time, a stable, useful team velocity.
How many teams use both together
A common pattern: use T-shirt sizes during early roadmap planning or initial backlog triage, when items are still rough and the goal is a general sense of relative scale across many ideas quickly. As specific items move closer to being worked on and get refined through backlog refinement sessions, re-estimate them using story points for the more precise planning sprint commitment requires. This two-stage approach matches the level of estimation precision to how well-understood the work actually is at each stage.
A worked example
A product team doing quarterly roadmap planning has 40 candidate backlog items to roughly assess. Using T-shirt sizing, they quickly triage all 40 in about an hour — sorting them into S, M, L, and XL buckets based on general team intuition, without deep discussion of any single item. This gives leadership a rough sense of overall roadmap scope. As the quarter progresses and specific items are refined for upcoming sprints, the team re-estimates each one using story points during backlog refinement — now with clear acceptance criteria and a shared understanding, they can distinguish more precisely between, say, two items that were both broadly "M" in the initial T-shirt sizing pass but turn out to be a 3 and an 8 respectively once properly understood.
Common mistakes when choosing between the two
- Using story points for very early, poorly understood items, creating false precision the team doesn't actually have enough information to support.
- Using T-shirt sizing for sprint-level planning, where the lack of precision makes it harder to build a reliable, useful team velocity over time.
- Converting T-shirt sizes directly to story points with a fixed formula (like "M always equals 5 points"), missing the genuine re-assessment that should happen once an item is better understood.
- Switching estimation methods inconsistently without team-wide agreement, creating confusion about which method applies when.
FAQ
Can T-shirt sizes be converted directly into story points? It's possible to create a rough mapping, but this should be treated as a starting point, not a fixed conversion — a genuine re-estimation once the item is better understood usually produces a more accurate story point value.
Which method is better for a brand-new agile team? T-shirt sizing can be a gentler introduction to relative estimation for a team new to agile, since it requires less precision and calibration than jumping straight into story points and Fibonacci scales.
Does T-shirt sizing support calculating team velocity? Not typically — velocity calculations depend on a consistent numeric scale (story points), which T-shirt sizing's broad categories don't provide in a form suitable for this kind of tracking.
How many T-shirt size categories should a team use? Most teams use 4-5 categories (XS, S, M, L, XL), which provides enough differentiation for rough triage without adding the complexity of a much finer-grained scale.